The Effect of Using Different Thermodynamic Models with Harmony Search Algorithm in the Accuracy of RNA Secondary Structure Prediction

Abstract

Ribonucleic acid (RNA) is a nucleic acid composed of a group of the nucleotides. RNA molecule is essential to all biological systems. The RNA strand folds back into itself during the folding process via hydrogen bonds to build the secondary and tertiary structures. Understanding the biological function of a given RNA molecule is critical to determining its structure. Since the structure of RNA molecules is a key to their function, algorithms for the prediction of RNA structure are promising. This paper discusses the effect of applying different thermodynamic models to HSRNAFold an RNA secondary structure prediction algorithm based on Harmony search (HS). The experiments were performed on twelve individual known structures from four RNA classes (5S rRNA, Group I intron 23S rRNA, Group I intron 16S rRNA and 16S rRNA). The data demonstrate that the results obtained via RNAeval are better than those of enf2 in terms of prediction accuracy. In addition, the time needed by RNAeval is less than the time needed by enf2 for the same number of iterations.